From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-10.2 required=3.0 tests=BAYES_00, HEADER_FROM_DIFFERENT_DOMAINS,MAILING_LIST_MULTI,MENTIONS_GIT_HOSTING, NICE_REPLY_A,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED,USER_AGENT_SANE_1 autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id BD002C5519F for ; Mon, 16 Nov 2020 14:45:31 +0000 (UTC) Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by mail.kernel.org (Postfix) with ESMTP id 1A7D322280 for ; Mon, 16 Nov 2020 14:45:31 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 1A7D322280 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=arm.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=owner-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix) id 97FD36B0070; Mon, 16 Nov 2020 09:45:30 -0500 (EST) Received: by kanga.kvack.org (Postfix, from userid 40) id 930E86B0071; Mon, 16 Nov 2020 09:45:30 -0500 (EST) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 7A8456B0072; Mon, 16 Nov 2020 09:45:30 -0500 (EST) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0058.hostedemail.com [216.40.44.58]) by kanga.kvack.org (Postfix) with ESMTP id 3900C6B0070 for ; Mon, 16 Nov 2020 09:45:30 -0500 (EST) Received: from smtpin22.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay01.hostedemail.com (Postfix) with ESMTP id D2473180AD807 for ; Mon, 16 Nov 2020 14:45:29 +0000 (UTC) X-FDA: 77490554778.22.stamp19_240200627329 Received: from filter.hostedemail.com (10.5.16.251.rfc1918.com [10.5.16.251]) by smtpin22.hostedemail.com (Postfix) with ESMTP id 9F12318038E67 for ; Mon, 16 Nov 2020 14:45:29 +0000 (UTC) X-HE-Tag: stamp19_240200627329 X-Filterd-Recvd-Size: 20849 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by imf38.hostedemail.com (Postfix) with ESMTP for ; Mon, 16 Nov 2020 14:45:28 +0000 (UTC) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 1D20431B; Mon, 16 Nov 2020 06:45:28 -0800 (PST) Received: from [10.37.12.42] (unknown [10.37.12.42]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPSA id 865573F718; Mon, 16 Nov 2020 06:45:25 -0800 (PST) Subject: Re: [PATCH mm v10 00/42] kasan: add hardware tag-based mode for arm64 To: Andrey Konovalov , Andrew Morton Cc: Catalin Marinas , Will Deacon , Dmitry Vyukov , Andrey Ryabinin , Alexander Potapenko , Marco Elver , Evgenii Stepanov , Branislav Rankov , Kevin Brodsky , kasan-dev@googlegroups.com, linux-arm-kernel@lists.infradead.org, linux-mm@kvack.org, linux-kernel@vger.kernel.org References: From: Vincenzo Frascino Message-ID: <5998e497-964e-3ef9-a98d-e45065c6f40a@arm.com> Date: Mon, 16 Nov 2020 14:48:32 +0000 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101 Thunderbird/68.10.0 MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset=utf-8 Content-Language: en-US Content-Transfer-Encoding: 7bit X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=1.2.4 Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: On 11/13/20 10:15 PM, Andrey Konovalov wrote: > This patchset adds a new hardware tag-based mode to KASAN [1]. The new mode > is similar to the existing software tag-based KASAN, but relies on arm64 > Memory Tagging Extension (MTE) [2] to perform memory and pointer tagging > (instead of shadow memory and compiler instrumentation). > > This patchset is co-developed and tested by > Vincenzo Frascino . > > This patchset is available here: > > https://github.com/xairy/linux/tree/up-kasan-mte-v10 > > For testing in QEMU hardware tag-based KASAN requires: > > 1. QEMU built from master [4] (use "-machine virt,mte=on -cpu max" arguments > to run). > 2. GCC version 10. > > [1] https://www.kernel.org/doc/html/latest/dev-tools/kasan.html > [2] https://community.arm.com/developer/ip-products/processors/b/processors-ip-blog/posts/enhancing-memory-safety > [3] git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux for-next/mte > [4] https://github.com/qemu/qemu > Tested-by: Vincenzo Frascino > ====== Overview > > The underlying ideas of the approach used by hardware tag-based KASAN are: > > 1. By relying on the Top Byte Ignore (TBI) arm64 CPU feature, pointer tags > are stored in the top byte of each kernel pointer. > > 2. With the Memory Tagging Extension (MTE) arm64 CPU feature, memory tags > for kernel memory allocations are stored in a dedicated memory not > accessible via normal instuctions. > > 3. On each memory allocation, a random tag is generated, embedded it into > the returned pointer, and the corresponding memory is tagged with the > same tag value. > > 4. With MTE the CPU performs a check on each memory access to make sure > that the pointer tag matches the memory tag. > > 5. On a tag mismatch the CPU generates a tag fault, and a KASAN report is > printed. > > Same as other KASAN modes, hardware tag-based KASAN is intended as a > debugging feature at this point. > > ====== Rationale > > There are two main reasons for this new hardware tag-based mode: > > 1. Previously implemented software tag-based KASAN is being successfully > used on dogfood testing devices due to its low memory overhead (as > initially planned). The new hardware mode keeps the same low memory > overhead, and is expected to have significantly lower performance > impact, due to the tag checks being performed by the hardware. > Therefore the new mode can be used as a better alternative in dogfood > testing for hardware that supports MTE. > > 2. The new mode lays the groundwork for the planned in-kernel MTE-based > memory corruption mitigation to be used in production. > > ====== Technical details > > From the implementation perspective, hardware tag-based KASAN is almost > identical to the software mode. The key difference is using MTE for > assigning and checking tags. > > Compared to the software mode, the hardware mode uses 4 bits per tag, as > dictated by MTE. Pointer tags are stored in bits [56:60), the top 4 bits > have the normal value 0xF. Having less distict tags increases the > probablity of false negatives (from ~1/256 to ~1/16) in certain cases. > > Only synchronous exceptions are set up and used by hardware tag-based KASAN. > > ====== Benchmarks > > Note: all measurements have been performed with software emulation of Memory > Tagging Extension, performance numbers for hardware tag-based KASAN on the > actual hardware are expected to be better. > > Boot time [1]: > * 2.8 sec for clean kernel > * 5.7 sec for hardware tag-based KASAN > * 11.8 sec for software tag-based KASAN > * 11.6 sec for generic KASAN > > Slab memory usage after boot [2]: > * 7.0 kb for clean kernel > * 9.7 kb for hardware tag-based KASAN > * 9.7 kb for software tag-based KASAN > * 41.3 kb for generic KASAN > > Measurements have been performed with: > * defconfig-based configs > * Manually built QEMU master > * QEMU arguments: -machine virt,mte=on -cpu max > * CONFIG_KASAN_STACK_ENABLE disabled > * CONFIG_KASAN_INLINE enabled > * clang-10 as the compiler and gcc-10 as the assembler > > [1] Time before the ext4 driver is initialized. > [2] Measured as `cat /proc/meminfo | grep Slab`. > > ====== Notes > > The cover letter for software tag-based KASAN patchset can be found here: > > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=0116523cfffa62aeb5aa3b85ce7419f3dae0c1b8 > > ====== History > > Change v9->v10: > (Andrey:) > - Rebase onto the mm tree. > - Drop "s390/kasan: include asm/page.h from asm/kasan.h" as linux-next > already contains this fix queued up with other s390 KASAN changes. > - Add kfence checks in mm/kasan/hw_tags.c. > - Rename KASAN_GRANULE_PAGE to KASAN_MEMORY_PER_SHADOW_PAGE. > - Clarify "slab memory" in documentation. > - Add a comment to CC_HAS_WORKING_NOSANITIZE_ADDRESS. > - Drop "kasan: kasan_non_canonical_hook only for software modes" as > KASAN_INLINE already depends on KASAN_GENERIC || KASAN_SW_TAGS. > - Mark kasan_init_hw_tags() as __init, as non-init parts have been moved > to kasan_init_hw_tags_cpu(). > - Rename kasan_(un)poison_memory() to (kasan_)(un)poison_range(). > (Vincenzo:) > - Add smb_wmb() to copy_highpage(). > - Rename mte_enable() to mte_enable_kernel(). > - Address review comments on selftest code. > - Address concerns from checkpatch.pl (courtesy of Marco Elver). > - Integrate a fixup from Marco Elver for code alignment. > > Changes v8->v9: > (Andrey:) > - Drop __init for kasan_init_hw_tags. > - Split out mte_enable() from mte_init_tags(). > - Fix incorrect commit author. > - Revert addr_has_metadata() change done in a previous version, otherwise > KASAN fails to print metadata for page_alloc allocations. > - Split kasan_init_hw_tags() into kasan_init_hw_tags_cpu() that is called > for each CPU in cpu_enable_mte(), and kasan_init_hw_tags() that is called > for boot CPU in smp_prepare_boot_cpu(). > - Move kunit_kasan_expectation struct definition under CONFIG_KASAN. > (Vincenzo:) > - Address Catalin's comments for "arm64: mte: Reset the page tag in > page->flags". > - New patch "kasan, mm: untag page address in free_reserved_area". > > Changes v7->v8: > (Andrey:) > - Rebased onto 5.10-rc2+. > - Enable in-kernel MTE via kasan_init_hw_tags() instead of doing it > directly from cpu_enable_mte(). This changes mte_init_tags() to both > init RRND and enable in-kernel MTE in sync mode. > - Put the patches back into a logical order. > - Rename KASAN_SHADOW_GRANULE_SIZE to KASAN_GRANULE_SIZE in > lib/test_kasan_module.c. > - Rename kasan_init_tags() to kasan_init_sw_tags() and mark as __init. > > Changes v6->v7: > (Andrey:) > - Rebase onto 5.10-rc2. > - Fix pgd_t not defined build warning on s390. > - Mark kasan_init_depth() as static. > - Don't use mte_*() functions directly in report_hw_tags.c > - Add hw_ prefix to KASAN-level definitions of arch_*() tag helpers. > - Add missing include to report_generic.h. > > Changes v5->v6: > (Vincenzo:) > - Re-based on 5.10-rc1. > - Modified the flow of the mte fault handler in order to address an issue > with word at a time routines that would affect Android init process. > - Dropped Reviewed-by from the involved patches. > (Andrey:) > - Properly use #if CONFIG_KASAN_STACK instead of #ifdef > CONFIG_KASAN_STACK_ENABLE. > - Expand CONFIG_KASAN checks in arch/arm64/kernel/kaslr.c and > arch/arm64/kernel/module.c. > - Don't select CONFIG_CONSTRUCTORS for HW_TAGS mode. > - Check PageSlab() in addr_has_metadata(). > > Changes v4->v5: > (Vincenzo:) > - Reset the tag associated by the kernel to a page when this is changed by > the user. > - Add a kselftest to verify that GCR_EL1 is preserved during context > switch. > - Squashed the TBI patch. > - Addressed some review comments. > - Verified the series with LTP. > (Andrey:) > - Put core arm64 patches first as requested by Catalin. > > Changes v3->v4: > (Vincenzo:) > - Introduced mte-def.h that contains MTE definitions. > - Moved __MTE_PREAMBLE in mte.h. > - Clarified which API is part of mte-kasan.h. > - Removed tsk argument from mte_set_kernel_gcr(). > - Addressed several nitpicks pointed out during review. > (Andrey:) > - Move include in include/linux/kasan.h to avoid build > failures. > - Don't move "select SLUB_DEBUG if SLUB" back and forth between patches > in KASAN Kconfig. > - Add arm64 prefix to "kasan: don't allow SW_TAGS with ARM64_MTE" commit. > - Don't add braces when defining KASAN_GRANULE_SIZE. > - Make KASAN_HW_TAGS compatible with init_on_alloc/free. > > Changes v2->v3: > (Vincenzo:) > - Use ARM64_ASM_PREAMBLE for asm macros. > - Rename mte-helper.h to mte-kasan.h. The new header is meant to contain > only macros and prototypes directly used in KASAN. The rest is defined > in mte.h. > - Update mte_get_mem_tag()/mte_get_random_tag() to used directly asm > volatile() macros instead of calling library functions. > - Harden mte_assign_mem_tag_range() to prevent an infinite loop in case of > unaligned size. > - Made sure that report_tag_fault() is executed only once. > - Simplify the mte code in __cpu_setup. > - Remove kprobes.h from mte.c includes. > - General cleanup of the code. > (Andrey:) > - Use READ/WRITE_ONCE when accessing reported in do_tag_recovery(). > - Move .unreq mte_tcr under CONFIG_ARM64_MTE to avoid build errors when MTE > is not enabled. > - Improve mm/kasan/shadow.c comment header. > - Clarify what is a memory granule in "kasan: rename KASAN_SHADOW_* to > KASAN_GRANULE_" commit description. > - Rename (report_)tags_sw/hw.c to to (report_)sw/hw_tags.c and drop > unnecessary rename commit. > - Adopt 100 lines limit for some mm/kasan/ changes. > - Align arguments for stack_trace_save() call in mm/slub.c. > - Restore comment before kasan_init_tags(). > - Remove GNU headers from all mm/kasan/ files. > - Simplify check_invalid_free() implementation tag-based modes. > - Drop subsequently removed report_tag_fault() implementation. > - Add KASAN_GRANULE_PAGE and use instead of PAGE_SIZE * KASAN_GRANULE_SIZE. > - Move kasan_enable/disable_current() declarations to simplify > include/linux/kasan.h. > - Drop dependency on CONFIG_SLUB_DEBUG. > - Clarify the purpose of CONFIG_STACKTRACE in KASAN Kconfig. > > Changes v1->v2: > - Rebase onto v10 of the user MTE patchset. > - Only enable in-kernel MTE when KASAN_HW_TAGS is enabled. > - Add a layer of arch-level indirection, so KASAN doesn't call MTE helpers > directly (this will be useful in case more architectures will add support > for HW_TAGS). > - Don't do arm64_skip_faulting_instruction() on MTE fault, disable MTE > instead. > - Don't allow software tags with MTE via arch/arm64/Kconfig instead of > lib/Kconfig.kasan. > - Rename mm/kasan/tags.c to tags_sw.c and mte.c to tags_hw.c, and do the > same for report_*.c files. > - Reword HW_TAGS Kconfig help text to make it less MTE specific. > - Reword and clarify Documentation. > - Drop unnecessary is_el1_mte_sync_tag_check_fault(). > - Change report_tag_fault() to only call kasan_report() once HW_TAGS is > introduced. > - Rename arch/arm64/include/asm/mte_asm.h to mte-helpers.h and move all > MTE-related defines and some helper functions there. > - Change mm/kasan/kasan.h to include mte-def.h instead of mte.h. > - Add WARN_ON() on unaligned size to mte_set_mem_tag_range(). > - Implement ldg/irg MTE routines as inline assembly. > - Remove smp_wmb() from mte_set_mem_tag_range(). > - Drop __must_check from mte_set_mem_tag_range() as KASAN has no use for > the return value. > - Drop zero size check from mte_assign_mem_tag_range(). > - Drop unnecessary include from low-level arm64 code. > - Move enabling TBI1 into __cpu_setup(). > - Drop stale comment about callee-saved register from > arch/arm64/kernel/entry.S. > - Mark gcr_kernel_excl as __ro_after_init. > - Use GENMASK() in mte_init_tags(). > > Andrey Konovalov (33): > kasan: drop unnecessary GPL text from comment headers > kasan: KASAN_VMALLOC depends on KASAN_GENERIC > kasan: group vmalloc code > kasan: shadow declarations only for software modes > kasan: rename (un)poison_shadow to (un)poison_range > kasan: rename KASAN_SHADOW_* to KASAN_GRANULE_* > kasan: only build init.c for software modes > kasan: split out shadow.c from common.c > kasan: define KASAN_MEMORY_PER_SHADOW_PAGE > kasan: rename report and tags files > kasan: don't duplicate config dependencies > kasan: hide invalid free check implementation > kasan: decode stack frame only with KASAN_STACK_ENABLE > kasan, arm64: only init shadow for software modes > kasan, arm64: only use kasan_depth for software modes > kasan, arm64: move initialization message > kasan, arm64: rename kasan_init_tags and mark as __init > kasan: rename addr_has_shadow to addr_has_metadata > kasan: rename print_shadow_for_address to print_memory_metadata > kasan: rename SHADOW layout macros to META > kasan: separate metadata_fetch_row for each mode > kasan, arm64: don't allow SW_TAGS with ARM64_MTE > kasan: introduce CONFIG_KASAN_HW_TAGS > arm64: kasan: Align allocations for HW_TAGS > arm64: kasan: Add arch layer for memory tagging helpers > kasan: define KASAN_GRANULE_SIZE for HW_TAGS > kasan, x86, s390: update undef CONFIG_KASAN > kasan, arm64: expand CONFIG_KASAN checks > kasan, arm64: implement HW_TAGS runtime > kasan, arm64: print report from tag fault handler > kasan, mm: reset tags when accessing metadata > kasan, arm64: enable CONFIG_KASAN_HW_TAGS > kasan: add documentation for hardware tag-based mode > > Vincenzo Frascino (9): > arm64: Enable armv8.5-a asm-arch option > arm64: mte: Add in-kernel MTE helpers > arm64: mte: Reset the page tag in page->flags > arm64: mte: Add in-kernel tag fault handler > arm64: kasan: Allow enabling in-kernel MTE > arm64: mte: Convert gcr_user into an exclude mask > arm64: mte: Switch GCR_EL1 in kernel entry and exit > kasan, mm: untag page address in free_reserved_area > kselftest/arm64: Check GCR_EL1 after context switch > > Documentation/dev-tools/kasan.rst | 82 ++- > arch/arm64/Kconfig | 9 +- > arch/arm64/Makefile | 7 +- > arch/arm64/include/asm/assembler.h | 2 +- > arch/arm64/include/asm/cache.h | 3 + > arch/arm64/include/asm/esr.h | 1 + > arch/arm64/include/asm/kasan.h | 5 +- > arch/arm64/include/asm/memory.h | 15 +- > arch/arm64/include/asm/mte-def.h | 14 + > arch/arm64/include/asm/mte-kasan.h | 67 ++ > arch/arm64/include/asm/mte.h | 22 +- > arch/arm64/include/asm/processor.h | 2 +- > arch/arm64/include/asm/string.h | 5 +- > arch/arm64/include/asm/uaccess.h | 23 + > arch/arm64/kernel/asm-offsets.c | 3 + > arch/arm64/kernel/cpufeature.c | 3 + > arch/arm64/kernel/entry.S | 41 ++ > arch/arm64/kernel/head.S | 2 +- > arch/arm64/kernel/hibernate.c | 5 + > arch/arm64/kernel/image-vars.h | 2 +- > arch/arm64/kernel/kaslr.c | 3 +- > arch/arm64/kernel/module.c | 6 +- > arch/arm64/kernel/mte.c | 118 +++- > arch/arm64/kernel/setup.c | 2 +- > arch/arm64/kernel/smp.c | 2 + > arch/arm64/lib/mte.S | 16 + > arch/arm64/mm/copypage.c | 9 + > arch/arm64/mm/fault.c | 59 ++ > arch/arm64/mm/kasan_init.c | 19 +- > arch/arm64/mm/mteswap.c | 9 + > arch/arm64/mm/proc.S | 23 +- > arch/arm64/mm/ptdump.c | 6 +- > arch/s390/boot/string.c | 1 + > arch/x86/boot/compressed/misc.h | 1 + > include/linux/kasan-checks.h | 2 +- > include/linux/kasan.h | 125 ++-- > include/linux/mm.h | 2 +- > include/linux/moduleloader.h | 3 +- > include/linux/page-flags-layout.h | 2 +- > include/linux/sched.h | 2 +- > include/linux/string.h | 2 +- > init/init_task.c | 2 +- > kernel/fork.c | 4 +- > lib/Kconfig.kasan | 65 +- > lib/test_kasan.c | 2 +- > lib/test_kasan_module.c | 2 +- > mm/kasan/Makefile | 25 +- > mm/kasan/common.c | 577 ++---------------- > mm/kasan/generic.c | 51 +- > mm/kasan/generic_report.c | 165 ----- > mm/kasan/hw_tags.c | 89 +++ > mm/kasan/init.c | 17 +- > mm/kasan/kasan.h | 74 ++- > mm/kasan/quarantine.c | 10 - > mm/kasan/report.c | 256 ++------ > mm/kasan/report_generic.c | 327 ++++++++++ > mm/kasan/report_hw_tags.c | 42 ++ > mm/kasan/{tags_report.c => report_sw_tags.c} | 14 +- > mm/kasan/shadow.c | 516 ++++++++++++++++ > mm/kasan/{tags.c => sw_tags.c} | 24 +- > mm/page_alloc.c | 9 +- > mm/page_poison.c | 2 +- > mm/ptdump.c | 13 +- > mm/slab_common.c | 2 +- > mm/slub.c | 29 +- > scripts/Makefile.lib | 2 + > tools/testing/selftests/arm64/mte/Makefile | 2 +- > .../arm64/mte/check_gcr_el1_cswitch.c | 155 +++++ > 68 files changed, 2036 insertions(+), 1165 deletions(-) > create mode 100644 arch/arm64/include/asm/mte-def.h > create mode 100644 arch/arm64/include/asm/mte-kasan.h > delete mode 100644 mm/kasan/generic_report.c > create mode 100644 mm/kasan/hw_tags.c > create mode 100644 mm/kasan/report_generic.c > create mode 100644 mm/kasan/report_hw_tags.c > rename mm/kasan/{tags_report.c => report_sw_tags.c} (87%) > create mode 100644 mm/kasan/shadow.c > rename mm/kasan/{tags.c => sw_tags.c} (92%) > create mode 100644 tools/testing/selftests/arm64/mte/check_gcr_el1_cswitch.c > -- Regards, Vincenzo